A cooperative surveillance system based on motion detection profile of car black-box

Abstract

As the number of surveillance cameras has been continuously increased, the number of surveillance camera for vehicle is steadily increasing. Due to the limit of view angle of camera, it is difficult to find the event that other car impacts on side or corner of own car. In this paper, we proposed the cooperative surveillance system based on profile of motion detection in car black-box, which extends detection coverage of car black-box. The proposed system consists of three layers: device layer which is car black-box, smart phone layer which connects device with cloud, and cloud layer which locates relative position of car black-box by analyzing profile of motion detection time instances. By using the cooperative surveillance system based on profile, if there is no event detection in user, the event detection probability could increases through the localization. Although the simulation was performed with one parking lot, event detection probability of cooperative surveillance system is increased. Only by using event detection time, parking time, leaving time, GPS information and floorplan, relative location could be estimated and relative location accuracy was 89.4% in our simulation environment.